Example Compliance Assurance Monitoring Submittals › ttnchie1 › mkb › documents › Supplement_1.pdffacilities and historical data obtained from the monitoring system. The development

1U.S. Environmental Protection Agency. Technical Guidance Document: ComplianceAssurance Monitoring, August 1998. Available on the EPA web site athttp://www.epa.gov/ttn/emc/cam.html.

CAM TECHNICAL GUIDANCE DOCUMENTAPPENDIX A

6/02 A-1

EXAMPLE COMPLIANCE ASSURANCE MONITORING SUBMITTALS

The purpose of this document is to supplement Appendix A of the Compliance AssuranceMonitoring (CAM) Technical Guidance1. The example CAM submittals presented in thissupplement are based upon “case studies” of the current monitoring approaches in use at actualfacilities and historical data obtained from the monitoring system. The development process forthese examples included: (1) identifying facilities which currently monitor control deviceparameters, had long-term monitoring data available for review, had conducted a performance/compliance test, and were willing to participate, (2) obtaining information on the monitoringapproach and monitoring data from the facility, (3) reviewing and analyzing the monitoringapproach and data, (4) discussing the information with plant personnel and, in some cases,conducting a site visit, and (5) preparing an example monitoring approach submittal from theinformation.

The basic approach used was to evaluate the monitoring conducted by the facility againstCAM general (design) and performance criteria. A monitoring approach submittal based uponthe facility’s current monitoring, modified as necessary to comply with CAM requirements, wasthen drafted. If sufficient information was available to evaluate alternative approaches (e.g.,different indicators, indicator ranges, or data averaging periods), alternative approaches alsowere investigated. Note that the resulting examples are not necessarily the only acceptablemonitoring approaches for the facility or similar facilities; they are simply examples ofapproaches used by particular facilities. The owner or operator of a similar facility may proposea different approach that satisfies part 64 requirements. Also, the permitting authority mayrequire additional monitoring.

One purpose of this supplement is to provide nonprescriptive examples of monitoringapproaches that meet the CAM submittal requirements for the specific cases studied. Eachexample monitoring submittal contains background information (including identification of thepollutant specific emissions unit), a description of the monitoring approach, and the rationale forselecting the indicators and indicator ranges. These examples represent the level of detailrecommended by EPA, but States may develop their own guidance as to the level of detail (moreor less) required in CAM monitoring approach submittals. Table 1 lists the examples containedin this supplement. Information has been collected for other control devices and monitoringapproaches and example monitoring approach submittals for these cases are being prepared forfuture release.

CAM TECHNICAL GUIDANCE DOCUMENTAPPENDIX A

A-2 6/02

Table 1. Example CAM Submittals Included in this SupplementNumber Example Title

A.4b Scrubber for VOC Control - Facility Q

A.9b Wet Electrostatic Precipitator (WESP) for PM Control - Facility P

A.11 Electrified Filter Bed (EFB) for PM Control - Facility K

A.16 Control Device Bypass - Facility R

A.17 Venturi Scrubber for PM Control - Facility S

A.18 Carbon Adsorber for VOC Control - Facility T

A.19a Baghouse for PM Control - Facility V

A.19b Baghouse for PM Control - Facility V

A.20 Absorber for SO2 Control - Facility W

A.24 Carbon Adsorber for VOC Control - Facility EE

A.25 Electrostatic Precipitator (ESP) for PM Control - Facility FF

A.27 Flue Gas Recirculation (FGR) for NOx Control - Facility HH

CAM TECHNICAL GUIDANCE DOCUMENTA.4B PACKED BED SCRUBBER FOR VOC CONTROL OF A BATCH PROCESS

6/02

A.4b PACKED BED SCRUBBER FOR VOC CONTROL OF A BATCH PROCESS – FACILITY Q

This page intentionally left blank


6/02 A.4b-1

EXAMPLE COMPLIANCE ASSURANCE MONITORING:PACKED BED SCRUBBER FOR VOC CONTROL – FACILITY Q

I. Background

A. Emissions Unit

Description: Batch mixers and tanks used in a chemicalprocess

Identification: Scrubber B-67-2

Facility: Facility QAnytown, USA

B. Applicable Regulation, Emissions Limit, and Monitoring Requirements

Regulation: Permit, State regulation

Emissions limit: VOC: 3.6 pounds per hour

Monitoring requirements: Inlet water flow, acetic acid concentration inscrubber underflow

C. Control Technology Packed bed scrubber

II. Monitoring Approach

The key elements of the monitoring approach for VOC are presented in Table A.4b-1. Theselected indicators of performance are the scrubber inlet water flow rate and the acetic acidconcentration in the scrubber water underflow. The scrubber inlet water flow rate is measuredcontinuously and recorded twice daily. The scrubber water underflow is sampled twice daily; the acetic acid concentration of each sample is determined by titration.


A.4b-2 6/02

TAB

LE A

.4b-

1. M

ON

ITO

RIN

G A

PPR

OA

CH

Indi

cato

r No.

1In

dica

tor N

o. 2

I.In

dica

tor

Mea

sure

men

t App

roac

h

Scru

bber

inle

t wat

er fl

ow ra

te.

Ace

tic a

cid

conc

entra

tion

in u

nder

flow

.

The

scru

bber

inle

t wat

er fl

ow ra

te is

mea

sure

d us

ing

a ra

diom

eter

.A

sam

ple

of th

e un

derf

low

is ta

ken

and

the

acet

ic a

cid

conc

entra

tion

dete

rmin

ed b

y tit

ratio

n.

II.

Indi

cato

r Ran

geA

n ex

curs

ion

is d

efin

ed a

s any

ope

ratin

g co

nditi

onw

here

the

scru

bber

inle

t wat

er fl

ow ra

te is

less

than

4 gp

m.

An

excu

rsio

n w

ill tr

igge

r an

inve

stig

atio

n of

the

occu

rren

ce, c

orre

ctiv

e ac

tion,

and

a re

porti

ngre

quire

men

t.

An

excu

rsio

n is

def

ined

as a

ny o

pera

ting

cond

ition

whe

re th

e un

derf

low

ace

tic a

cid

conc

entra

tion

isgr

eate

r tha

n 10

per

cent

. A

n ex

curs

ion

will

trig

ger a

nin

vest

igat

ion

of th

e oc

curr

ence

, cor

rect

ive

actio

n, a

nd a

repo

rting

requ

irem

ent.

III.

Perf

orm

ance

Crit

eria

A.

Dat

a R

epre

sent

ativ

enes

s

The

scru

bber

inle

t wat

er fl

ow ra

te is

mea

sure

d us

ing

a va

riabl

e ar

ea fl

ow m

eter

(rad

iom

eter

) loc

ated

in th

esc

rubb

er w

ater

inle

t lin

e. T

he m

inim

um a

ccep

tabl

eac

cura

cy o

f the

met

er is

±5

perc

ent o

f the

mea

sure

dva

lue

and

the

rang

e is

0 to

15

gpm

.

The

acet

ic a

cid

conc

entra

tion

in th

e sc

rubb

er w

ater

efflu

ent i

s mea

sure

d by

titra

ting

a w

ater

sam

ple

extra

cted

from

the

scru

bber

und

erflo

w.

B.

Ver

ifica

tion

ofO

pera

tiona

l Sta

tus

NA

NA

C.

Qua

lity

Ass

uran

ce a

ndC

ontro

l Pra

ctic

esA

nnua

l cal

ibra

tion

and

clea

ning

of r

adio

met

er.

Acc

epta

nce

crite

ria: ±

5 pe

rcen

t of t

he m

easu

red

valu

e.

Onl

y tra

ined

per

sonn

el p

erfo

rm sa

mpl

ing

and

titra

tion.

La

bora

tory

QA

/QC

pro

cedu

res a

re fo

llow

ed.

Cal

ibra

tion

stan

dard

s are

pre

pare

d to

ens

ure

the

sam

ple

titra

tion

is b

eing

per

form

ed a

ccur

atel

y.

D.

Mon

itorin

g Fr

eque

ncy

The

scru

bber

inle

t wat

er fl

ow ra

te is

mea

sure

dco

ntin

uous

ly a

nd re

cord

ed tw

ice

daily

. Th

e sc

rubb

er w

ater

out

let a

cetic

aci

d co

ncen

tratio

n is

mea

sure

d tw

ice

daily

.

Dat

a C

olle

ctio

n Pr

oced

ures

The

scru

bber

inle

t wat

er fl

ow ra

te is

reco

rded

twic

eda

ily.

(The

pos

t-con

trol e

mis

sion

s fro

m th

is u

nit a

rele

ss th

an th

e m

ajor

sour

ce th

resh

old,

so c

ontin

uous

mon

itorin

g an

d re

cord

ing

is n

ot re

quire

d.)

A w

ater

sam

ple

is ta

ken

and

titra

ted

man

ually

with

phen

olph

thal

ein

and

NaO

H so

lutio

n. (

The

post

-con

trol

emis

sion

s fro

m th

is u

nit a

re le

ss th

an th

e m

ajor

sour

ceth

resh

old,

so c

ontin

uous

mon

itorin

g an

d re

cord

ing

isno

t req

uire

d.)

Ave

ragi

ng P

erio

dN

one.

Non

e.


6/02 A.4b-3

MONITORING APPROACH JUSTIFICATION

I. Background

The pollutant specific emissions unit (PSEU) consists of process equipment in the celluloseesters division controlled by a packed bed scrubber. The process consists of batch mixers thatare used to convert cellulose into cellulose ester. Each mixer may be started at a different timeand may be used to make several batches per day. While in the mixers, the intermediate productis dissolved in acetic acid. The ester solution is transferred to storage tanks before being pumpedinto the next step in the process. A vent system collects the vapors from the mixers and tanksand a fan operated at constant speed pulls the vapors through the vent lines and into the scrubber. It is not possible for the gas to bypass the scrubber. The VOC load to the scrubbers in thisdivision primarily consists of acetic acid (and other carboxylic acids).

The scrubber is 4 feet in diameter and has about 8 feet of 2-inch packing. Fresh water issprayed at the top of the packing at 4 to 6 gpm; water from the underflow is recirculated to themiddle of the scrubber. The normal exit gas flow rate is approximately 1800 acfm.

II. Rationale for Selection of Performance Indicators

A packed bed scrubber is used to reduce VOC emissions from part of a chemicalmanufacturing process. Both batch mixers and process tanks are vented to this scrubber. Theprocesses in this area of the facility are mostly semi-batch operations, so the production rate atany one time varies. Therefore, it is difficult to relate the production rate to the VOC loadvented to this scrubber.

To comply with the applicable emission limit, a minimum water flow rate must be suppliedto the scrubber to absorb a given amount of VOC in the gas stream, given the size of the towerand height of the packed bed. The liquid to gas (L/G) ratio is a key operating parameter of thescrubber. If the L/G ratio decreases below the minimum, sufficient mass transfer of the pollutantfrom the gas phase to the liquid phase will not occur. The minimum liquid flow required tomaintain the proper L/G ratio at the maximum gas flow and vapor loading through the scrubbercan be determined. Maintaining this minimum liquid flow, even during periods of reduced gasflow, will help ensure that the required L/G ratio is achieved at all times. The concentration ofacetic acid in the scrubber underflow can be related to the water flow rate and acetic acidemissions, based on emissions test results and process modeling.

III. Rationale for Selection of Indicator Ranges

The indicator ranges were selected based on engineering calculations using ASPEN®process modeling software, emissions test data, and historical data. Computer modeling of thescrubber system was performed for the maximum allowable VOC concentration in the scrubberexhaust; the inlet water flow rate necessary for achieving adequate control was determined forseveral concentrations of acetic acid in the underflow. The scrubber efficiency was calculatedusing data obtained from emissions testing. The scrubber was modeled using an equilibrium-


A.4b-4 6/02

10

11

12

13

14

15

3 4 5 6

Water Feed (gpm)

Und

erflo

w C

onc.

(% a

cid)

Figure A.4b-1. Compliance curve.

based distillation method and ideal behavior of the gas phase was assumed; liquid phase activitycoefficients were estimated from a Wilson parameter fit of vapor-liquid equilibria data. It wasassumed that the control device delivers three actual stages of counter-current mass transfer witha recycle stream pumped from the effluent to the center of the column to ensure adequatedistribution of the liquid over the packing. The engineering model was calibrated for accuracyusing the results of source testing conducted while at normal operating conditions.

Figure A.4b-1 is a plot of the modeledoperating conditions (inlet water flow andscrubber underflow acetic acid concentration)necessary to maintain compliance. The linerepresents the operating conditions at maximumallowable emissions (3.6 lb VOC/hr); thescrubber’s VOC emissions are below the limitwhen the scrubber is operated at conditions thatfall below this line. For example, operating at ascrubber water flow rate of 4 gpm with an aceticacid concentration in the scrubber underflow of12 percent provides a margin of compliance withthe permitted VOC emission rate. The selectedindicator ranges for inlet water flow andunderflow acetic acid concentration were chosenbased on the compliance curve and normaloperating conditions. The indicator range(acceptable operating range) is defined as anyoperating condition where the scrubber inletwater flow is greater than 4 gpm and the scrubber underflow acetic acid concentration is lessthan 10 percent.

The 4 gpm level was chosen because it is the lower end of the preferred operating range. The 10 percent value was chosen because it is less than any point on the compliance curve (seeFigure A.4b-1), and the 1997 historical data show that all measured concentration data were lessthan 8.4 percent (typical values were between 2 and 6 percent). When an excursion occurs(scrubber inlet water flow of less than 4 gpm and/or scrubber underflow acetic acidconcentration of greater than 10 percent), corrective action will be initiated, beginning with anevaluation of the occurrence to determine the action required to correct the situation. Allexcursions will be documented and reported.

The scrubber typically operates at a water flow rate of 4 to 6 gpm. Figure A.4b-2 showsscrubber water flow data collected in 1997. The range for the 1997 data is 3 to 9.5 gpm; themean scrubber water flow rate was 5.3 gpm. There are four values less than 4 gpm, indicatingfour excursions. The bulk of the data falls between 5 and 6 gpm. Corrective action typically istaken (the flow is increased) when the scrubber water flow begins to fall below 5 gpm in order toavoid an excursion.


6/02 A.4b-5

0

1

2

3

4

5

6

7

8

9

10

0 200 400 600

Observation (2 per day)

Wat

er F

low

Rat

e, g

pm

Figure A.4b-2. 1997 scrubber water flow rate data.

0

1

2

3

4

5

6

7

8

9

0 100 200 300 400 500 600 700

Observation (2 per day)

Und

erflo

w A

cetic

Aci

d C

onc.

, %

Figure A.4b-3. 1997 underflow acetic acid concentration data.

Historical data from 1997 show the acetic acid concentration in the underflow is typicallyless than 6 percent. Figure A.4b-3 shows scrubber underflow acetic acid concentration data for1997. The maximum concentration was 8.4 percent, which is within the CAM indicator range. The mean concentration was 3.9 percent. The values decrease toward the end of the yearbecause production was decreased due totemporary changes in the market for a keyproduct. This further verifies thecorrelation between the acid concentra-tion in the underflow and the VOC load tothe scrubber. Because historical datashow that the scrubber routinely operateswithin the indicator range, there is notmuch variability in the data during typicalproduction periods, and the post-controlemissions from this scrubber are belowthe major source threshold, the water flowrate and acid concentration are recordedonly twice daily.

An emissions test was conducted onthis scrubber in December 1994. Anacetic acid sampling train validated usingEPA Method 301 was used to measureacetic acid emissions and EPA Methods 1through 4 were used to determine vent gas


A.4b-6 6/02

0

1

2

3

4

5

6

7

8

9

2 3 4 5 6 7 8 9 10

Water Feed, gpm

Und

erflo

w C

onc.

, % A

cetic

Aci

d

Figure A.4b-4. 1997 underflow acetic acid concentration vs. scrubber water flow.(2 measurements per day)

volumetric flow rates. The permitted emission limit is 3.6 lb VOC/hr. The average emissionsduring testing were 0.2 lb/hr, well below the emissions allowed for this scrubber. The inletwater flow rate was 5 gpm and the average scrubber underflow acetic acid concentration was5 percent. The test parameters and measured emissions and underflow concentration were usedin the ASPEN® computer model to calculate the efficiency of the scrubber. The model was thenused with that same efficiency to generate the compliance curve in Figure A.4b-1.

Figure A.4b-4 shows the underflow acetic acid concentration versus the scrubber waterflow rate for 1997. There were four excursions in 1997; the flow rate was less than 4 gpmduring those excursions, but the underflow acid concentration was always less than 10 percent.

CAM TECHNICAL GUIDANCE DOCUMENTA.9b WET ELECTROSTATIC PRECIPITATORS FOR PM CONTROL OF VENEER DRYERS

6/02

A.9b WET ELECTROSTATIC PRECIPITATORS (WESP) FOR PM CONTROL OF VENEER DRYERS – FACILITY P


6/02 A.9b-1

EXAMPLE COMPLIANCE ASSURANCE MONITORINGWET ELECTROSTATIC PRECIPITATORS (WESP) FOR PM CONTROL – FACILITY P

I. Background

A. Emissions Unit

Description: Steam-heated dryers used in plywoodmanufacturing

Identification: Veneer Dryers 1-6 (EU2)

APCD ID: WESP 1, WESP 2

Facility: Facility PAnytown, USA

B. Applicable Regulation and Emission Limit

Regulation No.: Permit, State Regulation

Emission limits: Particulate Matter (PM): 0.3 lb/1,000 ft2 (MSF) dried (3/8-inch thickness

basis)

Monitoring Requirements: Monitor WESP secondary voltage, quench inlettemperature, and WESP outlet temperature.

C. Control Technology Wet electrostatic precipitator


The key elements of the monitoring approach are presented in Table A.9b-1. The selectedindicators of performance are: WESP secondary voltage, quench inlet temperature, and WESPoutlet temperature. The selected indicator ranges are based on hourly average values.


A.9b-2 6/02

TAB

LE A

.9b-

1. M

ON

ITO

RIN

G A

PPR

OA

CH

Indi

cato

r No.

1In

dica

tor N

o. 2

Indi

cato

r No.

3

I.In

dica

tor

Mea

sure

men

t App

roac

h

WES

P se

cond

ary

volta

ge.

Que

nch

inle

t tem

pera

ture

.W

ESP

outle

t tem

pera

ture

.

The

WES

P se

cond

ary

volta

ge is

mon

itore

d us

ing

a vo

ltmet

er.

The

gas t

empe

ratu

re is

mea

sure

dw

ith a

ther

moc

oupl

e at

the

quen

chin

let.

The

gas t

empe

ratu

re is

mea

sure

d w

ith a

ther

moc

oupl

e at

the

WES

P ou

tlet.

II.

Indi

cato

r Ran

geA

n ex

curs

ion

is d

efin

ed a

s an

hour

lyav

erag

e vo

ltage

less

than

35

kV.

Excu

rsio

ns tr

igge

r an

inve

stig

atio

n,co

rrec

tive

actio

n, a

nd a

repo

rting

requ

irem

ent.

An

excu

rsio

n is

def

ined

as a

n ho

urly

aver

age

quen

ch in

let t

empe

ratu

re>3

75/F

. Ex

curs

ions

trig

ger a

nin

vest

igat

ion,

cor

rect

ive

actio

n, a

nda

repo

rting

requ

irem

ent.

An

excu

rsio

n is

def

ined

as a

n ho

urly

aver

age

outle

t tem

pera

ture

>17

5/F.

Ex

curs

ions

trig

ger a

n in

vest

igat

ion,

corr

ectiv

e ac

tion,

and

a re

porti

ngre

quire

men

t.

III.

Perf

orm

ance

Crit

eria

A.

Dat

a R

epre

sent

ativ

enes

sTh

e m

onito

ring

syst

em c

onsi

sts o

f avo

ltmet

er th

at is

par

t of t

he W

ESP

inst

rum

enta

tion

(TR

con

trolle

r).

The

min

imum

acc

urac

y of

the

voltm

eter

is ±

0.5

kV.

The

mon

itorin

g sy

stem

con

sist

s of a

ther

moc

oupl

e lo

cate

d in

the

quen

chin

let d

uctw

ork.

The

min

imum

accu

racy

of t

he th

erm

ocou

ple

is±2

.2/C

(±4/

F) o

r 0.7

5 pe

rcen

t of t

hem

easu

red

tem

pera

ture

in /C

,w

hich

ever

is g

reat

er.

The

mon

itorin

g sy

stem

con

sist

s of a

ther

moc

oupl

e lo

cate

d in

the

WES

P ou

tlet

duct

wor

k. T

he m

inim

um a

ccur

acy

of th

eth

erm

ocou

ple

is ±

2.2/

C (±

4/F)

or 0

.75

perc

ent o

f the

mea

sure

d te

mpe

ratu

re in

/C, w

hich

ever

is g

reat

er.

B.

Ver

ifica

tion

ofO

pera

tiona

l Sta

tus

NA

NA

NA

C.

QA

/QC

Pra

ctic

es a

ndC

riter

iaV

oltm

eter

zer

o ch

eck

durin

gsc

hedu

led

mai

nten

ance

per

form

edev

ery

3 w

eeks

.

Ther

moc

oupl

es c

alib

rate

d an

nual

lyby

com

paris

on a

gain

st a

n in

stru

men

tof

kno

wn

accu

racy

. Th

e ac

cept

ance

crite

ria is

±4/

F.

Ther

moc

oupl

es c

alib

rate

d an

nual

ly b

yco

mpa

rison

aga

inst

an

inst

rum

ent o

fkn

own

accu

racy

. Th

e ac

cept

ance

crit

eria

is ±

4/F.

D.

Mon

itorin

g Fr

eque

ncy

The

volta

ge o

n ea

ch W

ESP

ism

onito

red

cont

inuo

usly

(one

dat

apo

int p

er m

inut

e).

The

quen

ch in

let t

empe

ratu

re is

mon

itore

d co

ntin

uous

ly (o

ne d

ata

poin

t per

min

ute)

.

The

WES

P ou

tlet t

empe

ratu

re is

mon

itore

d co

ntin

uous

ly (o

ne d

ata

poin

tpe

r min

ute)

.

Dat

a C

olle

ctio

n Pr

oced

ure

Dat

a ar

e re

cord

ed o

n th

e co

ntin

uous

para

met

er m

onito

ring

syst

em(C

PMS)

com

pute

r.

Dat

a ar

e re

cord

ed o

n th

e C

PMS

com

pute

r.D

ata

are

reco

rded

on

the

CPM

S co

mpu

ter.

Ave

ragi

ng P

erio

dH

ourly

blo

ck a

vera

ge.

Hou

rly b

lock

ave

rage

.H

ourly

blo

ck a

vera

ge.


6/02 A.9b-3


I. Background

The pollutant-specific emissions units (PSEU) are the two WESPs that control six veneerdryers. The dryers are longitudinal, steam-heated dryers manufactured by Coe and Moore andare used in the manufacture of plywood. Veneer is introduced into the dryer either manually orusing automated veneer sheet feeders. The dried veneer sheets pass through a moisture detectoras they exit the dryer where any sheets not meeting moisture specifications are marked andsorted for redrying. Dry veneer sheets are coated with mixed glue and formed into panels.

Two WESPs, also referred to as E-tubes, remove particulate matter from the dryer exhaust. WESP No. 1 serves dryers Nos. 1, 5, and 6 and WESP No. 2 serves dryers Nos. 2, 3, and 4.


A WESP is designed to operate at a relatively constant voltage. A significant decrease involtage is indicative of a change in operating conditions that could lead to an increase inemissions. Low voltage can indicate electrical shorts or poor contacts that require maintenanceor repair of electrical components. However, the regular flush cycles the WESPs undergo toremove the particulate from the collection surfaces may also cause drops in voltage of shortduration. These brief voltage drops are part of the normal operation of the WESP.

Monitoring gas stream temperature can provide useful information about the performanceof a WESP. Quench inlet temperature primarily is an indication that the inlet gas stream is notso hot that a fire may develop in the duct work or WESP. In addition, the gas stream needs to becooled in order for some of the pollutants to condense. The WESP outlet temperature indicatesthat the gas stream has been sufficiently saturated to provide for efficient particle removal, andthat the water spray prior to the WESP inlet is functioning. High outlet temperatures could bethe result of plugged nozzles, malfunctioning pumps, or broken or plugged piping.


The selected indicator ranges are given below:

Secondary voltage: $35 kVQuench inlet temperature: #375°FStack outlet temperature: #175°F

An excursion is defined as (1) an hourly average voltage less than 35 kV; (2) an hourly averagequench inlet temperature greater than 375/F; or (3) an hourly average WESP outlet temperaturegreater than 175/F. When an excursion occurs, corrective action will be initiated beginning withan evaluation of the occurrence to determine the action required to correct the situation. Allexcursions will be documented and reported. An hourly average was chosen to account for theintermittent flush cycles the WESPs undergo that cause the voltage to drop temporarily.


A.9b-4 6/02

25

30

35

40

45

50

55

60

0 100 200 300 400 500 600 700

Observation

WES

P Vo

ltage

(kV)

Figure A.9b-1. October 1997 hourly averagesecondary voltage (WESP No. 1).

The indicator level for the WESP voltage was selected based upon the level maintainedduring normal operation. Typical operating voltages range from 35 to 55 kV. During the mostrecent performance test, the voltage ranged from 35 to 54 kV and the PM emissions were belowallowable levels. An indicator level at the low end of the normal operating range was selected(35 kV). During a malfunction (such as an electrical short), the WESP voltage levels areappreciably lower than normal operational levels. The voltage also drops for a short periodduring the normal flush cycles that are performed every few hours to clean the tube surfacewhere particulate is collected. Figure A.9b-1 displays the hourly average WESP secondaryvoltage during October 1997 for WESP No. 1.

The indicator levels for the quench inlet and WESP outlet gas temperatures also wereselected based on levels maintained during normal operation. High temperatures may indicate afire in the dryer or ductwork or a lack of water flow to the WESP. Temperature action levelswere selected that are slightly higher than normal operating temperatures. If the water flow tothe WESP is lost, the WESP outlet temperature will begin to approach the inlet temperature,which is much higher than 175/F. Figures A.9b-2 and A.9b-3 display the hourly average quenchinlet and WESP outlet temperature during October 1997 for WESP No. 1.


6/02 A.9b-5

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

Observation

Que

nch

Inle

t Tem

p (°

F)

Figure A.9b-2. October 1997 Hourly Average Quench Inlet Temperature(WESP No. 1)

0

20

40

60

80

100

120

140

160

0 100 200 300 400 500 600 700

Observation

Out

let T

empe

ratu

re (°

F)

Figure A.9b-3. October 1997 Hourly Average WESP Outlet Temperature(WESP No. 1)


A.9b-6 6/02

Indicator data for December 1995 to January 1996 and for October 1997 throughDecember 1997 were reviewed. These data included hourly average WESP secondary voltage,quench inlet temperature, and WESP outlet temperature measurements. The maximum hourlyaverage quench inlet temperature for WESP No. 1 was 336/F, while the maximum for WESPNo. 2 was 352/F. The maximum hourly average stack outlet temperature for WESP No. 1 was151/F, while the maximum stack outlet temperature for WESP No. 2 was 178/F. The averagemonthly voltages ranged from 47 to 51 kV for WESP No. 1 and from 40 to 46 kV for WESPNo. 2.

Data obtained during the most recent performance test (October 1996) confirmed the unitwas in compliance. During this test, the average measured PM emissions were 0.19 lb/MSFdried for WESP No. 1 and 0.21 lb/MSF dried for WESP No. 2. The measured particulateemissions were below the emission limitation of 0.3 lb/MSF dried (3/8-inch thickness basis). The WESP operating parameters during the performance test are summarized in Table A.9b-2.

TABLE A.9b-2. WESP OPERATING PARAMETERS DURING THE MOST RECENTPERFORMANCE TEST

WESPNo. Run

Production,ft2/hr

Particulate,lb/MSF dried

(3/8-inch basis)WESP voltage,

kVQuench inlet

T (/F)WESP outlet,

T (/F)

1 1 22,760 0.24 54 317 134

2 23,419 0.17 54 318 134

3 23,075 0.17 -- -- --

Average 23,085 0.19 54 318 134

2 1 23,899 0.24 35 328 147

2 32,238 0.17 38 332 143

3 26,897 0.20 40 331 147

Average 27,678 0.21 38 330 146

CAM TECHNICAL GUIDANCE DOCUMENTA.11 ELECTRIFIED FILTER BED FOR PM CONTROL OF VENEER DRYERS

6/02

A.11 ELECTRIFIED FILTER BED FOR PM CONTROL OF VENEER DRYERS – FACILITY K


6/02 A.11-1

EXAMPLE COMPLIANCE ASSURANCE MONITORINGELECTRIFIED FILTER BED (EFB) FOR PM CONTROL – FACILITY K

I. Background

A. Emissions Unit

Description: Natural gas-fired dryers used in plywoodmanufacturing

Identification: Veneer Dryer 1, Veneer Dryer 2

Facility: Facility KAnytown, USA

B. Applicable Regulation, Emission Limit, and Monitoring Requirements


Emission Limits: Particulate matter (PM): 0.30 lb/1000 ft2 (MSF) dried (3/8-inch thickness

basis), 4.1 lb/hr

Monitoring Requirements: EFB inlet temperature, EFB voltage, and EFBionizer current.

C. Control Technology EFB


The key elements of the monitoring approach are presented in Table A.11-1. The selectedindicators of performance are: EFB inlet temperature, voltage, and ionizer current. The selectedindicator ranges are based upon hourly average values.


A.11-2 6/02

TAB

LE A

.11-

1. M

ON

ITO

RIN

G A

PPR

OA

CH

Indi

cato

r No.

1In

dica

tor N

o. 2

Indi

cato

r No.

3

I.In

dica

tor

EFB

inle

t tem

pera

ture

. EF

B v

olta

ge.

EFB

ioni

zer c

urre

nt.

Mea

sure

men

t App

roac

hTe

mpe

ratu

re is

mea

sure

d us

ing

ath

erm

ocou

ple.

Vol

tage

is m

easu

red

with

avo

ltmet

er.

Ioni

zer c

urre

nt is

mea

sure

d w

ith a

nam

met

er.

II.

Indi

cato

r Ran

geA

n ex

curs

ion

is d

efin

ed a

s an

hour

lyav

erag

e EF

B in

let t

empe

ratu

re g

reat

erth

an 1

70/F

(>14

5/F

whe

n dr

ying

pin

eve

neer

). E

xcur

sion

s trig

ger a

nin

vest

igat

ion,

cor

rect

ive

actio

n, a

nd a

repo

rting

requ

irem

ent.

An

excu

rsio

n is

def

ined

as a

n ho

urly

aver

age

EFB

vol

tage

less

than

8 k

V.

Excu

rsio

ns tr

igge

r an

inve

stig

atio

n,co

rrec

tive

actio

n, a

nd a

repo

rting

requ

irem

ent.

An

excu

rsio

n is

def

ined

as a

n ho

urly

aver

age

EFB

ioni

zer c

urre

nt le

ssth

an 2

mA

. Ex

curs

ions

trig

ger a

nin

vest

igat

ion,

cor

rect

ive

actio

n, a

nda

repo

rting

requ

irem

ent.

III.

Perf

orm

ance

Crit

eria

A.

Dat

a R

epre

sent

ativ

enes

sTh

e m

onito

ring

syst

em c

onsi

sts o

f ath

erm

ocou

ple

inst

alle

d at

the

inle

t of

the

EFB

. Th

e m

inim

um a

ccur

acy

ofth

e th

erm

ocou

ple

is ±

2.2 /

C (±

4/F)

or

0.75

per

cent

of t

he m

easu

red

tem

pera

ture

in /C

, whi

chev

er is

grea

ter.

The

mon

itorin

g sy

stem

con

sist

s of a

voltm

eter

on

the

EFB

uni

t. T

hem

inim

um a

ccur

acy

of th

e vo

ltmet

eris

±0.

5 kV

.

The

mon

itorin

g sy

stem

con

sist

s of

an a

mm

eter

on

the

EFB

uni

t. T

hem

inim

um a

ccur

acy

of th

e am

met

eris

±0.

5 m

A.

B.

Ver

ifica

tion

ofO

pera

tiona

l Sta

tus

NA

NA

NA

C.

QA

/QC

Pra

ctic

es a

ndC

riter

iaTh

e ac

cura

cy o

f the

ther

moc

oupl

e is

chec

ked

annu

ally

(or a

s nee

ded)

by

calib

ratio

n us

ing

a si

gnal

tran

smitt

er.

The

ther

moc

oupl

e w

ells

are

perio

dica

lly c

heck

ed a

nd c

lean

ed (a

tle

ast a

nnua

lly).

Vol

tmet

er z

ero

is c

heck

ed w

hen

the

unit

is n

ot o

pera

ting.

Am

met

er z

ero

is c

heck

ed w

hen

the

unit

is n

ot o

pera

ting.

D.

Mon

itorin

g Fr

eque

ncy

The

EFB

inle

t tem

pera

ture

ism

easu

red

cont

inuo

usly

(at l

east

4 tim

es p

er h

our)

.

The

EFB

vol

tage

is m

easu

red

cont

inuo

usly

(at l

east

4 ti

mes

per

hour

).

The

EFB

ioni

zer c

urre

nt is

mea

sure

dco

ntin

uous

ly (a

t lea

st 4

tim

es p

erho

ur).

Dat

a C

olle

ctio

nPr

oced

ure

Dat

a ar

e st

ored

ele

ctro

nica

lly a

ndar

chiv

ed fo

r at l

east

5 y

ears

..D

ata

are

stor

ed e

lect

roni

cally

and

arch

ived

for a

t lea

st 5

yea

rs..

Dat

a ar

e st

ored

ele

ctro

nica

lly a

ndar

chiv

ed fo

r at l

east

5 y

ears

..

Ave

ragi

ng P

erio

dH

ourly

blo

ck a

vera

ge.

Hou

rly b

lock

ave

rage

.H

ourly

blo

ck a

vera

ge.


6/02 A.11-3


I. Background

The pollutant-specific emissions unit (PSEU) consists of two natural gas direct-fired veneerdryers controlled by an EFB. Dryer 1 is manufactured by Moore and has one zone and fourdecks. Dryer 2 is manufactured by Coe and has two zones and five decks. The dryers are usedin the manufacture of plywood.


Wood dryer exhaust streams contain dry PM, products of combustion and pyrolysis, andaerosols formed by the condensation of hydrocarbons volatilized from the wood chips. Sincesome of the pollutants from the dryers are in a gas phase at the normal dryer exhaust temperatureof 250/ to 300/F, these pollutants must be condensed in order to be collected by the EFB. Thegas stream is cooled to a temperature of about 180/F by the evaporative gas cooler that precedesthe EFB, using a water mist. The pollutants condense into fine liquid droplets and are carriedinto the EFB. The EFB ionizer gives the particles in the gas stream an electrical charge. Thehigh voltage electrode in the gravel bed creates charged regions on the gravel. As the gas passesthrough the bed, the charged particles are removed from the gas and transferred to the surface ofthe bed. Liquid and dust continuously build up on the gravel surface; the liquid slowly travelsthrough the bed and is allowed to drip into the drain outlet in the bottom of the unit. The gravelis periodically replaced (about one-third of the gravel is replaced each month).

Factors that affect emissions from wood dryers include wood species, dryer temperature,dryer residence time, dryer loading rate, and previous drying history of the wood. The rate ofhydrocarbon aerosol formation (from vaporizing the extractable portion of the wood) is lower atlower dryer temperatures. Small increases in dryer temperature can produce relatively largeincreases in the PM emission rate. If particles are held in the dryer too long, the surfaces canvolatilize; if these emissions are released into the ambient air, a visible blue haze can result.

The CAM indicators selected are EFB inlet temperature, EFB voltage, and EFB ionizercurrent. The EFB must be maintained at the proper temperature to allow collection of thehydrocarbon aerosol and particulate matter from the dryer. The EFB inlet temperature ismonitored to indicate the gas stream was cooled to the proper temperature range before enteringthe EFB and that the bed is operating at the proper temperature. Information from the EFBmanufacturer indicates that high EFB temperatures (e.g., temperatures in excess of 200/F) mayresult in excess stack opacity, as will low gravel levels (a low gravel level may cause insufficientPM collection). The voltage on the gravel and the current on the ionizer must be maintained sonegatively charged particles in the exhaust gas are attracted to positively charged regions on thegravel bed. An adequate ionizer current level indicates the corona is charging the particles in thegas stream. The bed voltage level indicates the intensity of the electric field in the bed. A dropin voltage or current could indicate a malfunction, such as a short or a buildup of dust orhydrocarbon glaze on the ionizer or the gravel. A short in the bed will show as high current withlittle or no voltage. A foreign object in the gravel bed which bridges the gap between the


A.11-4 6/02

electrode and grounded louvers can short the bed, as can a cracked electrical insulator. Thebed’s PM collection efficiency increases as the voltage and current increase within the unit’soperating range.

The parameters selected for monitoring are consistent with technical information on theoperation, maintenance, and emissions for EFB’s and dryers provided in EPA’s September 1992draft Alternative Control Technology (ACT) document for PM-10 emissions from the woodproducts industry. These parameters also were recommended by the manufacturer as parametersto monitor to ensure proper operation of the EFB unit.


Indicator data for June through August were collected and reviewed. These data includeEFB cooler inlet and outlet temperature, bed temperature, bed voltage, and ionizer currentmeasurements. No indicator ranges are specified in the current operating permit, but the permitdoes state that the EFB bed temperature shall not exceed 145/F when pine veneer is being dried. Based on the manufacturer’s recommendations, historical data, and data obtained during sourcetesting, the following indicator ranges were selected:

EFB bed inlet temperature: 2 mA

An excursion is defined as an hourly average of any parameter which is outside theindicator range. When an excursion occurs, corrective action will be initiated beginning with anevaluation of the occurrence to determine the action required to correct the situation. Allexcursions will be documented and reported.

Figure A.11-1 shows the hourly average EFB inlet temperature for June. The permitrequires that the EFB bed temperature be less than 145/F while drying pine veneer. The EFBinlet temperature is used as a surrogate for bed temperature. During normal operation, thetypical inlet temperature was 160 to 165/F when drying species other than pine. There wereshort periods of operation at 130 to 140/F when drying pine veneer, and lower temperatures thatindicate the dryers were not operating (e.g., on Fridays during the routine maintenanceshutdown). Similar operating ranges were observed for July and August. The maximum hourlyaverage EFB inlet temperatures for June, July, and August were 174/F, 173/F, and 176/F,respectively. The manufacturer recommends maintaining the EFB at a temperature of 160 to180/F. Therefore, based on this recommendation and on normal operating conditions, theindicator range chosen was an hourly average inlet temperature less than 170/F (less than 145/Fwhen drying pine veneer). If the EFB inlet temperature exceeds 170/F (145/F when dryingpine), corrective action will be initiated.

Figure A.11-2 shows the hourly average EFB voltage for June. From Figure A.11-2, it canbe observed that the EFB typically operates in the range of 10 to 15 kV. Some short periods of


6/02 A.11-5

MonthMean hourly average

current, mA

June 2.8

July 2

August 2

Average 2.3

MonthMean hourly average

voltage, kV

June 12.4

July 11.6

August 10.9

Average 11.6

operation occur from 5 to 10 kV. The mean hourly voltages for June, July, and August are givenbelow. These statistics do not include data from periods during which the EFB was notoperating and the voltage was recorded as 1.0 or zero. (For example, the EFB is shut downevery Friday for maintenance.)

The manufacturer’s recommended bed voltage range is 5 to 10 kV. The average voltagesduring the 1992, 1993, and 1996 performance tests were 6.7 kV, 11 kV, and 14 kV, respectively. Based on all data reviewed, greater than 8 kV was chosen as the indicator range for the hourlyaverage EFB bed voltage. If the hourly average bed voltage drops below 8 kV during periods ofnormal operation (excludes shutdown periods), corrective action will be initiated.

Figure A.11-3 shows the hourly average EFB ionizer current for the month of June. FromFigure A.11-3 it can be seen that the EFB typically operates at an ionizer current in the range of2 to 5 mA. The mean hourly average currents for June, July, and August are shown below. Inaddition, the manufacturer’s recommended range is 2 to 4 mA. Therefore, the indicator rangechosen was an hourly average current greater than 2 mA. If the hourly average ionizer currentdrops below 2 mA during normal operation (excludes shutdown periods), corrective action willbe initiated.

Emissions test results and indicator data are presented below for the 1992, 1993, and 1996performance tests. The 1992 and 1993 tests were conducted while drying pine; the 1996 test wasconducted while drying Douglas fir. The EFB is subject to a PM emission limitation of0.30 lb/MSF (4.1 lb/hr). Both limits were met during all three performance tests.


A.11-6 6/02

0

20

40

60

80

100

120

140

160

180

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Day

Tem

pera

ture

, °F

EFB taken offline once per week for maintenance

Figure A.11-1. June EFB inlet temperature (hourly average).

Year

PMemissions,

gr/dscf

PMemissions,

lb/MSF

PMemissions,

lb/hrAverage voltage,

kVAverage ionizer

current, mA

Average EFBinlet

temperature, /F

1992 0.016 0.16 1.5 6.7 4.9 153

1993 0.015 0.22 2.0 10.8 2.8 154

1996 0.02 0.30 1.1 14 1.4 189


6/02 A.11-7

0

2

4

6

8

10

12

14

16

18

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Da y

Volta

ge, k

V

Figure A.11-2. June EFB bed voltage (hourly average).

0

1

2

3

4

5

6

7

8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Day

Cur

rent

, mA

Figure A.11-3. June EFB ionizer current (hourly average).

This page intentionally left blank.

CAM TECHNICAL GUIDANCE DOCUMENTA.16 CONTROL DEVICE (BOILER) BYPASS

6/02

A.16 CONTROL DEVICE (BOILER) BYPASS – FACILITY R


6/02 A.16-1

EXAMPLE COMPLIANCE ASSURANCE MONITORINGCONTROL DEVICE (BOILER) BYPASS – FACILITY R

I. Background

A. Emissions Unit

Description: APCD (boiler) bypass valve

Identification: East and West boilers

Facility: Facility RAnytown, USA

B. Applicable Regulation, Emissions Limit, and Bypass Monitoring Requirements


Emissions Limits: CO: 200 ppm

Monitoring Requirements: Temperature downstream of bypass valve.

C. Control Device

Two boilers in parallel.


The key elements of the bypass monitoring approach are presented in Table A.16-1. Theselected indicators are the temperatures in the horizontal and vertical portions of the bypass linedownstream of the boiler bypass valve. The temperatures are measured continuously;instantaneous temperature values are recorded every 15 minutes.

Note: This compliance assurance monitoring example is presented as an illustration of oneapproach to monitoring for control device bypass. The example presents only theparameters monitored to ensure the control device is not being bypassed. Parameters toensure the control device is operating properly also are monitored, but are not discussedin this example.


A.16-2 6/02

TABLE A.16-1. BYPASS MONITORING APPROACHI. Indicator Vertical and horizontal bypass line temperatures

Measurement Approach Thermocouples downstream of bypass valve.

II. Indicator Range An excursion is defined as a vertical line temperature ofgreater than 550/F or a horizontal line temperature of greaterthan 250/F. An excursion shall trigger an inspection,corrective action as necessary, and a reporting requirement.

III. Performance CriteriaA. Data Representativeness

Gas temperature is measured using thermocouples in twolocations downstream of the bypass valve, prior to thecommon exhaust stack. The minimum accuracy of thethermocouples is 2.2/C (±4/F) or ±0.75 percent of thetemperature measured in /C, whichever is greater.

B. Verification of Operational Status NA

C. QA/QC Practices and Criteria The thermocouples are checked annually with a redundanttemperature sensor. Acceptance criteria: ±15/F of themeasured value.

D. Monitoring Frequency The temperatures are measured and recorded every15 minutes.

Data Collection Procedures The temperatures are recorded by the computer controlsystem every 15 minutes.

Averaging period None.


6/02 A.16-3

Figure A.16-1. Process schematic.


I. Background

The FCCU regenerator flue gas contains approximately 10 percent CO by volume, and isreferred to as “CO gas.” The CO gas is routed to two tangentially-fired boilers (East and West)in parallel, designed with sufficient residence time, turbulence, and temperature to fully combustthe CO to CO2. The exhaust from each boiler enters a common stack, where an emission limit of200 ppm CO must be met. The FCCU regenerator is equipped with piping that enables the COgas to bypass the boilers and flow directly to the common stack. Use of the bypass line isessential for the safe operation of the boilers during startup and shutdown periods. The piping isequipped with a butterfly valve. The position of this valve is monitored by the computer controlsystem, and is kept fully closed during normal operation. The operators routinely pack the valvewith ceramic fiber insulation to prevent leaks. A process schematic is shown in Figure A.16-1.


A.16-4 6/02

II. Rationale for Selection of Performance Indicator

Although the bypass valve position is computer-controlled, it has a tendency to leak if nottightly packed with insulation. Therefore, the operators need an indicator to detect leakage ofthe valve that might cause excess CO emissions. Testing was performed to determine the effectof boiler load on CO emissions. The results showed the boilers emitted negligible CO regardlessof operating load. The effect of a leaky valve on CO emissions (measured in the stack) and thegas temperature downstream of the bypass valve then was examined. The results showed that asthe amount of valve leakage increases and the CO concentration in the common stack increases,the temperature downstream of the valve also increases because of the high temperature of theCO gas (the temperature of the CO gas upstream of the valve is approximately 960/F). Therefore, the selected indicator of a leaky or open bypass valve is the temperature downstreamof the bypass valve.

III. Rationale for Selection of Indicator Range

A test program was conducted to determine the relationship between the gas temperaturedownstream of the bypass valve and the CO emissions. The gas temperature in the bypass lineand the CO concentration in the common stack were measured at baseline conditions (noleakage) and for eight different leak conditions. Temperature was measured at two locations: thevertical section of the bypass line (19 feet downstream of the valve) and the horizontal section ofthe bypass line (47 feet downstream of the valve). During normal conditions, when the CO levelin the common stack was less than 50 ppm, the temperature in the vertical section was roughly410/F, while the temperature in the horizontal section was 110/F.

To induce leakage of the valve, the valve was opened 5 percent on day 1 and 3 percent onday 2, and immediately closed. The packing material broke loose during each opening. Oninducing the leaks, the temperature downstream of the valve rose quickly and eventually reacheda stable temperature. To evaluate the effect of adding packing to the valve on downstreamtemperatures and CO levels in the common stack, the valve was progressively packed withceramic fiber insulation and allowed to stabilize. The level of CO in the stack and thedownstream temperatures decreased with the amount of insulation added.

For each of the seven test runs or conditions, multiple data points were collected andrecorded for the temperatures and the CO concentrations. Rather than calculating the average as the representative value for each run as is traditionally done with performance test data, apercentile measure was determined from the data for each run. The percentile value fortemperature and for CO concentration were selected independently. All of the temperaturereadings for the run were ranked from lowest to highest, and the value that coincides with the5th percentile for all of the temperature readings for that run was selected. Then, all of the COconcentration readings for the run were ranked lowest to highest, and the value that coincideswith the 95th percentile for all of the CO concentration readings for that run was selected. Thesepercentile values were selected to represent the test run instead of an average value. Table A.16-2 shows a summary of the readings for each test condition or run; both the average values and


6/02 A.16-5

the percentile values are shown. Table A.16-2 shows data for the vertical duct temperature,horizontal duct temperature, and CO concentration for each test condition.

Figures A.16-2 and A.16-3 show the relationship between CO emissions and the gastemperature at the horizontal and vertical locations. The 5th percentile temperature readingsreflect levels at the lower end of the range for each condition that can alert the boiler operator tobypass valve leakage. Conversely, since the CO levels varied during each test condition, the95th percentile CO levels for each test condition were selected to be conservative (on the highside). For added confidence, indicator ranges were developed for both measurement locations (itis expected that the two thermocouples will not fail at the same time). Based on the datacollected during testing, an excursion is defined as a vertical duct temperature of greater than550/F or a horizontal duct temperature of greater than 250/F. An excursion will trigger aninspection, corrective action as necessary, and a reporting requirement.

CA

M T

ECH

NIC

AL G

UID

AN

CE D

OC

UM

ENT

A.16 C

ON

TR

OL

DE

VIC

E (B

OIL

ER) B

YPA

SSA

.16-66/02

TABLE A.16-2. SUMMARY OF TEMPERATURE AND CO EMISSIONS LEVELS DURING TEST CONDITIONS

ConditionTest Period(minutes)

Vertical Temperature Readings(°F)

Horizontal TemperatureReadings (°F)

CO Level (ppmvd at 50%excess air)

Average 5th Percentile Average 5th Percentile Average 95th Percentile

Baseline -- Normal operation, minimal leakage 222 410 405 112 109 39.5 44.5

Open1 -- Open/close bypass valve to force leakage(day 2)

8 Transient Data Period

Leak -- Monitoring period following valveopen/close

98 683 641 463 426 351 358

Pack1 -- Monitoring period after one tube ofpacking was injected into valve


Pack2 -- Monitoring period after a second tube ofpacking was injected

57 676 671 453 449 229 230

Pack3 -- Monitoring period after a third tube ofpacking was injected

1084 634 629 341 307 169 191

Pack 45 -- Monitoring period after a fourth andfifth tube of packing was injected

176 482 443 179 160 30.0 35.7

Open 2 -- Close/open bypass valve to force leakagea second time (day 3)


Leak 2 -- Monitoring period following valveopen/close #2

105 641 604 443 411 242 248

Pack1X -- Monitoring period after one tube ofpacking was injected into valve after Leak 2


Pack 2X -- Monitoring period after a second tubeof packing was injected into valve after Leak2

122 588 577 397 389 123 127


6/02 A.16-7

Figure A.16-2. CO Level (95th Percentile) in the Common Stack vs. Horizontal Temperature Measurement (5th Percentile).

Figure A.16-3. CO Level (95th Percentile) in the Common Stack vs. Vertical Temperature Measurement (5th Percentile).

CAM TECHNICAL GUIDANCE DOCUMENTA.17 VENTURI SCRUBBER FOR PM CONTROL

6/02

A.17 VENTURI SCRUBBER FOR PM CONTROL--FACILITY S


6/02 A.17-1

EXAMPLE COMPLIANCE ASSURANCE MONITORINGVENTURI SCRUBBER FOR PM CONTROL: FACILITY S

I. Background

A. Emissions Unit

Description: Wood-fired boilerIdentification: Boiler AFacility: Facility S

Anytown, USA

B. Applicable Regulation, Emissions Limit, and Monitoring Requirements

Regulation: State regulation (Federally enforceable)

Emissions Limit: Particulate Matter (PM): Determined using the following equation:

P = 0.5 *(10/R)0.5

where:P = allowable weight of emissions of fly ash and/or other PM in

lb/mmBtu.

R = heat input of fuel-burning equipment in mmBtu/hr based onthe measured percent of O2 and volumetric flow rate.

The State rule also specifies that the opacity of visible emissions cannot beequal to or greater than 20 percent, except for one 6-minute period perhour of not more than 27 percent.

Monitoring Requirements: Continuous Opacity Monitoring System (COMS)

C. Control Technology

Venturi scrubber


The key elements of the monitoring approach are presented in Table A.17-1. Theindicators of performance are the boiler exhaust O2 concentration (a measure of excess air level)and the differential pressure across the scrubber venturi.

TABLE A.17-1. MONITORING APPROACHIndicator No. 1 Indicator No. 2

I. Indicator Exhaust gas oxygen concentration Scrubber differential pressure

Measurement Approach O2 monitor Differential pressure transducer.

II. Indicator Range An excursion is defined as an hourly boilerexhaust O2 concentration of less than 11 orgreater than 16 percent. Excursions triggeran inspection, corrective action, and areporting requirement.

An excursion is defined as a 1-hour averagedifferential pressure below 10.0 inches ofwater. Excursions trigger an inspection,corrective action, and a reportingrequirement.

III. Performance CriteriaA. Data Representativeness

The O2 monitor is located in the boilerexhaust.

The differential pressure transducermonitors the static pressures upstream anddownstream of the scrubber’s venturithroat.

B. Verification of Operational Status NA NA

C. QA/QC Practices and Criteria Daily zero and span checks. Adjust whendrift exceeds 0.5 percent O2.

Quarterly comparison to a U-tubemanometer. Acceptance criteria is0.5 in. w.c.

D. Monitoring Frequency Measured continuously. Measured continuously.

Data Collection Procedures 1-minute averages are computed anddisplayed. The PC then computes and storesa 1-hour average using the 1-minuteaverages.

1-minute averages are computed anddisplayed. The PC then computes andstores a 1-hour average using the 1-minuteaverages.

Averaging period 1-hour. 1-hour.

CA

M T

ECH

NIC

AL G

UID

AN

CE D

OC

UM

ENT

A.17 V

EN

TU

RI S

CR

UB

BE

R FO

R PM

CO

NT

RO

LA

.17-26/02


6/02 A.17-3


I. Background

The pollutant-specific emissions unit (PSEU) is PM from a wood-fired boiler. Particulatematter in the boiler’s exhaust stream is controlled by a venturi scrubber. A COMS is required bythe applicable State rule. However, water droplets in the boiler exhaust will interfere with theCOMS measurements and consequently make the use of a COMS impractical. An alternativemonitoring program utilizing parametric monitoring has been proposed. The monitoringapproach includes continuous monitoring of the wood-fired boiler’s excess air, the steamproduction rate, and the differential pressure across the scrubber’s venturi throat.


The operating conditions for this type of source (wood-fired boiler) can have a significantimpact on the amount of particulate emissions created. Furthermore, for a venturi scrubber, theinlet particulate matter loading to the scrubber will have an impact on the emissions level fromthe scrubber (i.e., emissions from the scrubber are expected to increase as the loading to thescrubber increases for the same scrubber operating conditions). Site-specific emissions test dataconfirm these expectations. Therefore, indicators of performance of both the control device andprocess were selected for this source.

The scrubber differential pressure was selected as the indicator of control deviceperformance. The differential pressure is proportional to the water flow and air flow through thescrubber venturi throat and is an indicator of the energy across the scrubber and the properoperation of the scrubber within established conditions.

Excess air levels can have a significant impact on boiler performance. Excess air isdefined as that air exceeding the theoretical amount necessary for combustion. Insufficientexcess air will result in incomplete combustion and an increase in emissions. A minimum ofabout 50 percent excess air is necessary for combustion of wood or bark fuels. Provision of toomuch excess air causes the furnace to cool and also can result in incomplete combustion. Therefore, the proper excess air level is important for proper operation of the boiler. The percentoxygen in the exhaust gas stream is an indicator of the excess air level (0 percent oxygen wouldequal 0 percent excess air, 8 percent oxygen is approximately 50 percent excess air, and12 percent oxygen is approximately 100 percent excess air).


A.17-4 6/02


Baseline information on the relationship among process operating conditions, controldevice operating conditions, and emissions was necessary to establish the indicators and ranges. A series of test runs was performed at several different boiler operating conditions becauseparametric monitoring is being proposed as an alternative to COMS.

Emissions tests were performed to establish a basis for indicator ranges that correspond tocompliance with the PM emissions limit. A set of nine test runs was performed on the boiler atthree different levels of steam generation (three test runs were performed at each steamgeneration level). Emissions sampling was based on EPA Methods 1 through 5 (40 CFR 60,Appendix A). The results of the first series of emissions tests indicated a problem meeting theemissions limits at the lower load level; the lack of a means to control excess air levels duringboiler operation was suspected as the cause of the excess emissions. A second series of testswere performed a year later after automatic boiler control equipment was installed. The secondseries of tests also was comprised of nine runs at three operating loads. The results of these18 tests were used in selecting the indicator ranges. The results of these tests are presented anddiscussed in the following paragraphs.

Figure 1 graphically presents the excess air level versus the nominal boiler load (steamgeneration rate) for the tests. During the first series of tests, before automatic boiler controlswere added, the boiler operated at a very high level of excess air (over 500 percent) at the low-level operating load, at a high level of excess air (over 200 percent) at the mid level operatingload, and below 200 percent at the high-level operating load. Without the automatic boilercontrols, the same amount of air was being introduced to the boiler regardless of the operatingload (wood feed rate), resulting in a significant increase in excess air levels as wood feed ratedecreased. After the automatic controls were added, the excess air was maintained at lowerlevels for the low-level and mid-level load conditions (less than 300 percent and 200 percent,respectively).

The results of the two test series are summarized in Table A.17-2. Three test runs wereperformed at each steam generation rate.


6/02 A.17-5

TABLE A.17-2. TEST RESULTSa

Nominal steamgeneration rate

(lb/hr)Venturi differentialpressure (in. H2O)

Boilerexhaust O2

(%)

Particulateemissions

(lb/MMBtu)

Allowableparticulateemissions

(lb/MMBtu)

Series 1:(Before BoilerControlModifications)

25,000 15.6 18.1 0.73 0.25

40,000 22.9 16.2 0.43 0.21

60,000 22.2 12.6 0.06 0.16

Series 2:(After BoilerControlModifications)

33,000 12.0 15.5 0.07 0.25

52,000 12.1 13.9 0.06 0.21

77,000 12.0 13.0 0.05 0.17a All values are 3-run averages.

At the first level of steam generation (25,000 lb/hr), the amount of excess air ranged from544 percent to 752 percent by volume. The particulate emissions rate ranged from 0.528 to1.12 lb/MMBtu. The maximum allowable emissions ranged from 0.23 to 0.27 lb/MMBtu. Themaximum allowable emissions varies because it is based on the heat input rate. The allowableemissions rate was exceeded for all three test runs. The second set of test runs was performed ata nominal steam generation level of 40,000 lb/hr. The amount of excess air ranged from 244 to830 percent. The particulate emissions rate ranged from 0.21 to 0.82 lb/MMBtu. The maximumallowable emissions ranged from 0.17 to 0.28 lb/MMBtu. The maximum allowable emissionsrate was exceeded for all three test runs. The third set of test runs was operated at a nominalsteam generation level of 60,000 lb/hr. The steam generation level actually ranged from60,000-70,000 lb/hr but dropped below 50,000 lb/hr midway through the third of the three testsperformed. The amount of excess air for these three test runs ranged from 123 to 188 percent. The particulate emissions rate ranged from 0.05 to 0.06 lb/MMBtu. The maximum allowableemissions ranged from 0.15 to 0.17 lb/MMBtu. The boiler was well within the maximumallowable emissions rate for all three test runs.

For the test series conducted after the addition of automatic controls, at the first level ofsteam generation (33,000 lb/hr nominal), the amount of excess air ranged from 255 to341 percent by volume (15 to 16 percent oxygen). The particulate emissions rate ranged from0.062 to 0.081 lb/MMBtu. The maximum allowable emissions ranged from 0.23 to0.29 lb/MMBtu. The particulate emissions were less than the allowable emissions rate for allthree test runs. The second set of test runs was performed at a nominal steam generation level of77,000 lb/hr. The amount of excess air ranged from 128 to 194 percent (12 to 14 percentoxygen). The particulate emissions rate ranged from 0.045 to 0.057 lb/MMBtu. The maximumallowable emissions ranged from 0.16 to 0.18 lb/MMBtu. The particulate emissions were lessthan the allowable emissions rate for all three test runs. The third set of test runs was performedat a nominal steam generation level of 52,000 lb/hr. The amount of excess air for these three testruns ranged from 196 to 223 percent (13 to 14 percent oxygen). The particulate emissions rateranged from 0.056 to 0.067 lb/MMBtu. The maximum allowable emissions ranged from 0.20 to


A.17-6 6/02

0.21 lb/MMBtu. The boiler operated within the maximum allowable emissions rate for all threetest runs.

Figure 2 presents the particulate emissions rate versus boiler load for the two test series. Figures 3 and 4 present the particulate emissions rate versus excess air and boiler exhaustoxygen level, respectively. The test results show that during the first test series the emissionsincrease significantly as the excess air increases. The allowable emissions limit was exceeded atthe low- and mid-level operating loads. The results of the second test series conducted afterautomatic boiler controls were added also show a relationship among the excess air level, boilerload, and particulate emissions rates. However, the particulate emissions rates were well withinthe allowable emissions rates for all test runs at all load conditions. Note that the performance ofthe system (boiler and venturi scrubber) was significantly better during the second series of testswhen the automatic boiler controls were being used to control air levels even though the venturiscrubber was operating at a lower pressure drop (12 versus 22 in. w.c.).

The indicator selected for monitoring boiler operation is exhaust gas oxygen concentration.The selected indicator range for the boiler exhaust gas oxygen is greater than 12 and less than16 percent O2 (one-hour average). The indicator range was chosen based upon the 1-hr test runaverages for the January 1999 test data. During these tests, the average oxygen concentrationwas maintained between 12 and 16 percent. The oxygen concentration is measuredcontinuously. An excursion triggers an inspection, corrective action, and a reportingrequirement. The selected range will promote maximum efficiency and provide a reasonableassurance that the boiler is operating normally.

The indicator range selected for monitoring venturi scrubber operation is a pressuredifferential of greater than 10 in. w.c. (one-hour average). An excursion triggers an inspection,corrective action, and a reporting requirement. The differential pressure is measured severaltimes per minute. A one-minute average is calculated, and an hourly average is calculated fromthe one-minute averages. The selected indicator range was chosen by examining theJanuary 1999 test data. During these tests, the differential pressure was maintained between 10and 15 in. w.c. The measured particulate emissions limit during these tests at all three boilerloads was approximately one third of the allowable emissions rate (large margin of compliance). Therefore, a differential pressure of greater than 10 in. w.c. was selected as the indicator range.


6/02 A.17-7

Figure 2: Particulate Emissions vs. Steam Flow Rate

0

0.2

0.4

0.6

0.8

1

1.2

0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000

Steam Flow Rate (lb/hr)

Part

icul

ate

Emis

sion

s (lb

/MM

Btu

)

Before Modification

After Modification

Figure 1: Excess Air vs. Steam Flow Rate

0

200

400

600

800

1000

0 20,000 40,000 60,000 80,000 100,000

Steam Flow Rate (lb/hr)

Exce

ss A

ir (P

erce

nt)

Before ModificationAfter Modification


A.17-8 6/02

Figure 4: Particulate Emissions vs. Exhaust Oxygen Level

0

0.2

0.4

0.6

0.8

1

1.2

10 11 12 13 14 15 16 17 18 19 20

Exhaust Oxygen Level (Percent)

Part

icul

ate

Emis

sion

s (lb

/MM

Btu

)

Before Modification

After Modification

Figure 3: Particulate Emissions vs. Excess Air

0

0.2

0.4

0.6

0.8

1

1.2

0 100 200 300 400 500 600 700 800 900Excess Air (Percent)

Part

icul

ate

Emis

sion

s (lb

/MM

Btu

)

Before Modification

After Modification

CAM TECHNICAL GUIDANCE DOCUMENTA.18 CARBON ADSORBER FOR VOC CONTROL

6/02

A.18 CARBON ADSORBER FOR VOC CONTROL – FACILITY T


6/02 A.18-1

EXAMPLE COMPLIANCE ASSURANCE MONITORINGCARBON ADSORBER FOR VOC CONTROL – FACILITY T

I. Background

A. Emissions Unit

Description: Loading Rack

Identification: LR-1

APCD ID: SRU-1

Facility: Facility TAnytown, USA

B. Applicable Regulation, Emission Limit, and Monitoring Requirements

Regulation: Permit

Emission Limits: VOC: 0.67 lb/1,000 gallons transferred

(80 mg/L transferred)

Monitoring Requirements: Monitor carbon adsorber outlet VOCconcentration, monitor position of APCDbypass valve, conduct a leak detection andrepair program.

C. Control Technology:

Carbon adsorber.


The key elements of the monitoring approach are presented in Table A.18-1. The carbonadsorber outlet VOC concentration in percent by volume as propane is continuously monitored. The selected indicator range is based on a 1-hour rolling average concentration. Periodic leakchecks of the vapor recovery unit also are conducted and the position of the carbon adsorberbypass valve is monitored to ensure bypass of the control device is not occurring.

Note: Facility T also monitors parameters related to the vapor tightness of connections and tanktrucks and other parameters of the vapor recovery system, but this example focuses on themonitoring performed on the carbon adsorber.

CA

M T

ECH

NIC

AL G

UID

AN

CE D

OC

UM

ENT

A.18 C

AR

BO

N A

DSO

RB

ER

FOR

VO

C C

ON

TR

OL

A.18-2

6/02

TABLE A.18-1. MONITORING APPROACHIndicator No. 1 Indicator No. 2

I. Indicator Outlet VOC concentration (percent). Equipment leaks.

Measurement Approach Breakthrough detector (NDIR analyzer). Monthly leak check of vapor recovery system.

II. Indicator Range An excursion is defined as an hourly average outlet VOCconcentration of 4 percent by volume (as propane) or greater. When this level is reached or exceeded, the loading rack willbe shut down via an automated interlock system. Anexcursion will trigger an investigation, corrective action, and areporting requirement.

An excursion is defined as detection of a leakgreater than or equal to 10,000 ppm (as methane)during normal loading operations. An excursionwill trigger an investigation, corrective action, and areporting requirement. Leaks will be repairedwithin 15 days.

III. Performance CriteriaA. Data

Representativeness

The analyzer is located at the carbon adsorber outlet. A handheld monitor is used to check for leaks in thevapor collection system during loading operations.

B. Verification ofOperational Status

NA NA

C. QA/QC Practices andCriteria

Daily zero/span drift. Adjust if drift is greater than 2.5 percentof span.

Follow procedures in 40 CFR 60, Appendix A,Method 21.

D. MonitoringFrequency

The outlet VOC concentration is monitored every 2 minutes. Monthly.

Data CollectionProcedures

The data acquisition system (DAS) collects the outlet VOCconcentration every 2 minutes and calculates a rolling 1-houraverage. Periods when breakthrough is detected and theinterlock system shuts down the loading rack also arerecorded.

Records of inspections, leaks found, leaks repaired.

Averaging period 1 hour (rolling). None.

APCD Bypass Monitoring: A pressure gauge on the vapor header line is used to detect if the relief valve is open. The valve opens if the pressurereaches 18 inches H2O. The DAS records the instantaneous pressure reading every 2 minutes.


6/02 A.18-3

E KV CL

=×× 106


I. Background

The pollutant specific emissions unit (PSEU) is a vacuum regenerative carbon adsorberused to reduce VOC emissions from a gasoline loading rack. (Note: This facility is not a majorsource of HAP emissions and is not subject to 40 CFR 63, Subpart R, or 40 CFR 60,Subpart XX.) The maximum throughput of the loading rack is 43,000,000 gallons per month,and the facility operates 24 hours per day, 7 days per week.

The carbon adsorber has two identical beds, one adsorbing while the other is desorbing ona 15-minute cycle. Carbon bed regeneration is accomplished with a combination of high vacuumand purge air stripping which removes previously adsorbed gasoline vapor from the carbon andrestores the carbon's ability to adsorb vapor during the next cycle. The vacuum pump extractsconcentrated gasoline vapor from the carbon bed and discharges into a separator. Non-condensed gasoline vapor plus gasoline condensate flow from the separator to an absorbercolumn which functions as the recovery device for the system. In the absorber, the hydrocarbonvapor flows up through the absorber packing where it is liquefied and subsequently recovered byabsorption. Gasoline product from a storage tank is used as the absorbent fluid. The recoveredproduct is simply returned along with the circulating gasoline back to the product storage tank Asmall stream of air and residual vapor exits the top of the absorber column and is recycled to theon-stream carbon bed where the residual hydrocarbon vapor is re-adsorbed.


A non-dispersive infrared (NDIR) analyzer is used to monitor the carbon adsorber outletVOC concentration in percent by volume as propane and ensure breakthrough is not occurring. This monitor provides a direct indicator of compliance with the VOC limit since it continuouslymeasures the outlet VOC concentration in percent. An interlock system is used to shut downloading operations when an excursion occurs.

A monthly leak inspection program also is performed to ensure that the vapors releasedduring loading are captured and conveyed to the vapor recovery unit. A handheld monitor isused to detect leaks in the vapor collection system. The position of the vapor recovery unit’srelief valve is monitored to ensure the control device is not bypassed.


The indicator range for the breakthrough detector was selected based on engineeringcalculations. The VOC emission rate can be expressed as follows (see 40 CFR 60.503):


A.18-4 6/02

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

0:00:00 4:00:00 8:00:00 12:00:00 16:00:00 20:00:00 0:00:00

Time

Con

cent

ratio

n, p

erce

nt (a

s pr

opan

e)

2-min

hourlyavg.

Figure A.18-1. A typical day’s concentration data.

where:E = emission rate of VOC, mg/LV = volume of air/vapor mixture exhausted, scmC = concentration of VOC, ppmL = volume loaded, LK = density of calibration gas, 1.83x106 mg/scm for propane

Assuming 100 percent displacement of all vapors into the vapor recovery unit (e.g., if300,000 L are loaded, 300,000 L of vapor pass through the unit) and assuming that breakthroughis occurring, it may be conservatively assumed that V is equal to L (V is actually less than L ifthe carbon adsorber is operating properly). Converting the volume displaced/exhausted(300,000 L) to cubic meters (300 scm) and substituting 300 scm for V, 80 mg/L for E, and1.83x106 mg/scm for K gives C equal to 43,700 ppm, or 4.4 percent. Therefore, the indicatorrange for the outlet VOC concentration is 4 percent (rolling hourly average), to provide areasonable assurance of compliance with the VOC limit of 80 mg/L loaded. If the hourlyaverage outlet VOC concentration reaches or exceeds 4 percent, the unit will be shut down andloading prevented via an automated interlock system. All excursions will be documented andreported. Figure A.18-1 presents both 2-minute instantaneous (dotted line) and hourly average(solid line) outlet VOC concentration data for a typical day’s operation. The outlet VOCconcentration typically is less than 0.5 percent as propane.

Documents

Example Compliance Assurance Monitoring Submittals › ttnchie1 › mkb › documents › Supplement_1.pdffacilities and historical data obtained from the monitoring system. The development